The Accountable Bureaucrat
abstract. Common wisdom has it that bureaucrats are unaccountable to the people they regulate and must therefore be closely supervised by elected officials or (perhaps ironically) the federal courts. For many detractors of the administrative state, as well as many proponents, agency accountability hangs on the concentrated power of the President in particular. This Article presents a different vision. Drawing on in-depth interviews with officials from numerous agencies, we show that everyday administrative practices and relationships themselves support accountability of a kind that neither elections nor judicial review alone can achieve.
Our interviews reveal that agency officials work within structures that promote the very values accountability is supposed to serve: deliberation, inclusivity, and responsiveness. Three primary features of the administrative state support this vision of accountability. First, political appointees and career civil servants, often presented as adversaries, actually represent complementary decision-making modalities. Appointees do not impose direct presidential control but imbue agencies with a diffuse, differentiated sense of abstract political values and policy priorities tied to the electoral and civil-society coalitions that support the administration in power. Civil servants use expertise and experience to set the parameters within which decisions can be made. Combining these different but interdependent approaches to policymaking promotes deliberation informed by public opinion and the public interest. Second, agencies work through what we refer to as a decision-making web, which facilitates continual justification and negotiation among officials with different roles inside the state. This claim stands in stark contrast to the strict hierarchy often attributed to government bureaucracy. We show how the principal-agent model gives way, more often than not, to the dispersion of decision-making power, which promotes the pluralistic inclusivity of views in a way hierarchical decision-making does not. Finally, numerous practices connect agencies directly and pervasively to the people and situations they regulate. Those required by law, like notice-and-comment rulemaking, are supplemented by varied other means by which agencies respond and adapt to the views of affected publics and the realities of the regulated world.
Our research provides crucial empirical evidence of how the everyday work of government gets done and gives the often-invoked notion of accountability some real content. It leads us to reject formalistic claims about what constitutes accountability in the abstract and to focus instead on the relationships, structures, and practices that actually promote accountability—features of the administrative state that help head off arbitrariness, incorporate multiple perspectives, and encourage negotiated, provisional outcomes. These resources for promoting republican democracy within the bureaucracy, however, are neither inherent nor eternal: they must be actively nourished. This Article thus should change how we think about government accountability and inform how we structure our institutions to achieve it.
authors. Anya Bernstein is Professor of Law, University of Connecticut School of Law. Substantial parts of this Article were written while she was a Visiting Scholar at the Max Planck Institute for Social Anthropology’s Department of Law and Anthropology. She thanks the Department and its director, Marie-Claire Foblets, for providing a supportive environment for intellectual pursuits, and Larissa Vetters for many helpful conversations. She is grateful for the generous support of the Fulbright Foundation, which supported that visit with a U.S. Fulbright Scholar award. Cristina Rodríguez is Leighton Homer Surbeck Professor of Law, Yale Law School. She is grateful for generous financial support from the Oscar M. Ruebhausen Fund at Yale Law School, which greatly assisted in the research supporting this Article. Both authors are deeply grateful for the excellent insights and feedback they received in response to this Article. They would like to thank Daniel Carpenter, Brian Cook, Brian Feinstein, David Fontana, Chris Havasy, Christine Jolls, Amy Kapczynski, Joshua Macey, Neysun Mahboubi, Jerry Mashaw, Alex Mechanick, Gillian Metzger, Jennifer Nou, Anne Joseph O’Connell, Nick Parrillo, Shalev Roisman, Roberta Romano, Susan Rose-Ackerman, Kate Shaw, Jodi Short, Reva Siegel, Norman Silber, David Spence, Glen Staszewski, Jed Stiglitz, John Witt, Ilan Wurman, and David Zaring, as well as participants in faculty workshops at Yale Law School, the University of Toronto Faculty of Law, the University of Connecticut Law School, and the Max Planck Institute for Social Anthropology, and those in the seminar on Power in the Administrative State. Special thanks to Errol Meidinger for encouraging this project in its earliest stages. This work also benefitted enormously from several generations of talented students who played essential roles in the conceptualization and execution of this multiyear project. For their dedication to the data gathering and coding phases of our project, the authors thank Ashraf Ahmed, Natasha Khan, Alexandra Lichtenstein, Alexander Mechanick, William Powell, Kshithij Shrinath, Becca Steinberg, and Helen White. For their excellent work in helping them design the project, they thank Anna Mohan and Megan McGlynn. And for their assistance in polishing this Article, they thank Colin Burke, David Hopen, Jim Huang, Beatrice Pollard, Chelsea Thomeer, Bardia Vaseghi, and Charlotte Witherspoon. The authors’ greatest thanks go to the numerous agency officials who generously shared their experiences and insights, not to mention their time, to make this research possible.
Introduction
The image of the unelected bureaucrat indicts the administrative state. Judges and critics regularly marshal this menacing figure to challenge the excessive power and questionable legitimacy of the bureaucracy as a whole. On one view, the unelected bureaucrat heightens the dangers posed by executive government. When the Supreme Court enjoined the Biden Administration’s regulation requiring many employees to be vaccinated or test regularly for COVID-19, Justice Gorsuch praised the decision for helping “to prevent ‘government by bureaucracy supplanting government by the people.’”1 The unelected bureaucrat also threatens the rightful power of the President. Chief Justice Roberts, on behalf of Court majorities, has expressed fear that removal protections for certain officials can lead agencies to “slip from the Executive’s control, and thus from that of the people.”2 Whether aggrandizing the executive branch or sapping its one elected official of power, the figure of the unelected bureaucrat is a reliably winning card.3 Even many who support robust administration sometimes accept the premise that unelected bureaucrats pose a threat to accountable government. The dissent in the vaccine-or-test case, for example, argued that the agency’s rule had “the virtue of political accountability, for [the Occupational Safety and Health Administration (OSHA)] is responsible to the President, and the President is responsible to—and can be held to account by—the American public.”4
These claims all rest on the commonsense notion that elections create government accountability.5 In this Article, we argue instead that accountability within the American administrative state arises from elections only indirectly, if at all. Our empirical research, involving interviews with administrators across a range of federal agencies,6 reveals numerous structures, relationships, and practices within the state itself that produce various and important forms of accountability. These scaffolds augment and complement the accountability created by elections, in some cases producing the very sorts of accountability that elections supposedly, but do not actually, provide. Based on these findings, we argue that placing excessive emphasis on elections or elected officials—whether to constrict, justify, or structure administration—gets in the way of understanding how to build and sustain an accountable democratic state.
Democracy depends on accountability: those who exercise power should be held responsible for their actions and decisions.7 At its most basic, accountability requires government actors to justify their positions so that others can evaluate, challenge, or override them.8 Such justification has distinct payoffs. It renders arbitrary or biased views more visible and contestable and pushes government actors to consider multiple perspectives in their decision-making. It allows interested publics to test the quality of government decisions and to change them over time, giving democratic subjects an ongoing role in their own governance.9 And it ensures that public institutions serve as sites for the contestations, negotiations, and provisional outcomes that characterize any successful democracy.10 This understanding of accountability fits with the notion of republican democracy, which rejects the arbitrary exercise of power or power that fails to take relevant interests into account.11 In short, accountability occurs when government is made deliberative, inclusive, and responsive.
Agency action is often regarded by judges and commentators as accountable only insofar as it can be directly controlled by the elected President.12 But our research locates the production of accountability elsewhere. We identify three mutually supporting aspects of agency practice as particularly important to ensuring accountability: (1) diffuse, rather than concentrated, forms of political control; (2) nonhierarchical organizational structures of negotiation and deliberation among numerous actors and groups; and (3) practices that keep agencies attuned to affected publics and events in the regulated world. In this Article, we present each of these features of administration and explain how they should change the way scholars and lawyers conceptualize and pursue accountability in government.
In Part I, we show that, as a matter of practice, direct presidential involvement in agency policymaking is neither frequent nor the most important source of political control. Instead, political influence is usually broader and more diffuse.13 Political appointees chosen by the President or the heads of agencies play a crucial role, but they do not merely carry out the President’s agenda. These appointees are usually relatively independent, acting on policy orientations congruent with, but not in any clear sense directed by, the President. Their bigger contribution lies in introducing forms of reasoning and decision-making that complement the work of career officials. The complex relationships between these epistemic communities help structure agency accountability. Meanwhile, presidential administration, though now entrenched, is often less significant than either its proponents or its detractors claim.14 We show how the broader political connection we identify helps ground accountability not through a direct connection to elections but through interdependent relationships between different modalities of decision-making.
In Part II, we bring to light the networked structures through which agency policymaking proceeds. Commentators sometimes assume that accountability in the administrative state rests on clear hierarchies and fairly simple principal-agent relationships.15 We found agency action to be characterized by something quite different: broad participation, multifarious input, and ongoing reason-giving characterized as much by negotiation as by supervision.16 Participants in this process continually justify their positions not just to a particular principal but to many players within the bureaucracy. And supervisors’ specific policy preferences often do not precede but rather emerge through policymaking practices themselves. Our findings thus challenge the simplified principal-agent model of agency practice and suggest that accountability can emerge from dispersed, rather than consolidated, authority.
In Part III, we pan out to consider agencies’ external relations. Agencies frequently and intentionally react to events in the world and the public opinion that arises from those events. Officials utilize both the formal channels created through the notice-and-comment process and myriad informal, semistructured channels they themselves have opened to engage with the public.17 In so doing, agencies render themselves subject to evaluation and influence by those whom their decisions affect. Bureaucrats’ relative insulation from elections leads some critics to assume that they are walled off from the world—computers in sealed rooms. But lacking a direct electoral connection does not keep administrators removed from regulated entities or the regulated world. Our findings suggest that, on the contrary, agencies have more diverse, frequent, and interactive relationships with the publics and situations they regulate than elections could provide.18 These ongoing interactions and agencies’ attendant concerns with the efficacy of their actions promote accountability as well.
We conclude by considering the implications of the relational, negotiated, and contextual forms of accountability we identify. A popular challenge to the administrative state paints it as undemocratic—in the most sinister formulation, a “deep” state that threatens our freedoms.19 We find, in contrast, that everyday agency processes facilitate accountability. They build in requirements for officials to continually present ideas to be reviewed, vetted, and tested by other actors who bring to bear different forms of judgment and expertise and who themselves hold meaningful stakes in the policymaking process. Such interactions routinely lead to the reconsideration of views with respect to legal questions, implementation issues, and policy desirability. These practices provide scaffolding that makes it possible for participants to enact the virtues of accountability: to be pragmatically responsive to social needs, to problem-solve in the public interest, and to justify the exercise of government power. The features of agency decision-making that we uncover are thus worth pursuing as a matter of institutional design, and their potential presence should inform both legal doctrines and political perceptions related to bureaucracy.
Elections, of course, remain crucial to legitimating
government,20 and their effects do and should
permeate agency decision-making.21 But real accountability requires
more.22
Narrowing the notion of accountability to the electoral connection instantiates
a peculiarly anemic notion of democracy that leaves out many of the traits that
make democratic governance normatively attractive.23 Doing so amounts to insisting that
elections do work for which they are not suited and ignores most of the
accountability-forcing work that needs to be done. The legitimation produced
through elections finds its complement in the practices we detail here, which
build accountability into government work itself. In some sense, then, the
imprecation that “unelected bureaucrats” are unaccountable is a complaint about
the wrong thing: if bureaucrats are unaccountable, it is not because they are
unelected but because they do not utilize the accountability structures that
administration provides.24
Unfortunately, the Supreme Court has begun to severely constrain the administrative state in the name of this anemic conception of democracy. Suspicion of the unelected bureaucrat is used to justify limiting both how Congress can design the regulatory state25 and how agencies can address urgent problems in their statutory purview.26 The Justices support these limits in the name of democracy, but their reliance on a formalistic conception of democratic legitimacy, unmoored from actual practices, threatens to undermine, rather than strengthen, agency accountability.
Our analysis also departs from a common assumption that government accountability depends largely on strict hierarchy or principal-agent relationships. The simplified principal-agent model, the foundation of much doctrine and legal scholarship on the administrative state, cannot account for the wealth of mutual accounting we found cutting across institutional hierarchies: the peers who needed to justify their positions to one another, the political appointees who had to persuade career staff to move an idea forward. These relationships, even when strictly speaking hierarchical, often do not function on a command-and-control basis. The model also fails to capture the dynamic development not just of policies but of policy preferences. It is often not the case that a superior commands a subordinate to carry out a predetermined policy position. Such a position is just as likely something that develops in the policymaking process itself: deliberation shapes preferences.27 Of course, agencies provide many opportunities for superiors to give clear instructions that subordinates must obey. But taking the simplified principal-agent relationship as the model for government action obscures more than it reveals.
In addition to challenging some of the conventions of administrative law, our research contributes to a growing field of empirical studies of bureaucracy, which to date has focused largely on the “street-level bureaucrats” who make individualized determinations in a given policy context.28 We study instead those we might call elevator-level administrators—the ones who produce the policies to begin with.29 We do so not to explain how any particular policy was produced or how any specific agency functions, but rather to illuminate policymaking trends and mechanisms broadly operative in the administrative state.30 Given the variety of agency forms and cultures within the American bureaucracy, we do not claim that our picture is comprehensive. But our interviewees, dispersed across different roles in various agencies, overlapped sufficiently in their descriptions of agency practices to provide important empirical insight into the relationship between administrative policymaking and a well-functioning democracy.31
We do not claim that administrative agencies are accountable in some transhistorical, inherent way. Accountability inevitably depends on empirical realities that differ across circumstances: institutions can be more or less, and also differently, accountable depending on their participants, their structures, their cultures, and so on.32 Instead of asking whether an institution is accountable, we ask what scaffolds it provides to support and channel accountability. Such scaffolds do not themselves create accountability. There is always the possibility that people will shirk, undermine, or attack them.33 But these scaffolds encourage and enable accountability. Those we found in our work rest upon an assumption of pluralism, provide means for mediating differing interests, and build responsiveness to outside interests and events into agency action. Again, these findings do not mean that the bureaucracy will necessarily be accountable; but they do mean that it has developed tools for accountability—tools that we should understand, defend, and even replicate.34
While we do not claim to present a complete or eternal picture of government accountability, our study raises normative and political questions that discussions of accountability too often ignore: questions about why accountability matters, what purposes it serves, and what it should look like in practice. That is, we not only provide empirical insight into accountability practices but advance the theorization of what accountability entails. We show that the bureaucracy can be accountable even though the officials who comprise it are unelected. But we also resist overclaiming the democratic bona fides and efficacy of presidential control. Our research should thus change how we think about accountability in the state and inform how we structure institutions to achieve it.
* * *
Before we get to our findings, a word on our methodology.35 Much of the information and insight of this project comes from thirty-nine in-depth, open-ended interviews with political appointees and career civil servants from eleven agencies, small and large, old and new, benefit-managing and conduct-regulating. Interviews are useful for eliciting new information and unknown views because they solicit participant perspectives, experiences, and ideas. Our primary objective in conducting the interviews was to better understand agency statutory interpretation and policymaking, and we therefore inquired into interpretive approaches and tools; relations to Congress, the President, other agencies, and states; judicial influence; and participants’ views on how agencies work with statutes. We asked no questions about accountability. Our conclusion that the practices interviewees described promoted government accountability emerged from our engagement with the interview transcripts, not from our subjects’ self-descriptions.
In devising our interview protocols, we did not aim to produce structurally identical, easily comparable responses; we sought instead complex, nuanced images of administration from the bureaucrat’s point of view.36 To find resonances in these disparate conversations, we identified major categories of analysis in the interview transcripts and coded each passage with its relevant categories.37 This engagement with the material also yielded an ambient sense of how our subjects’ descriptions converged, which further guided our approach to the data.38 Data collection, coding, and interpretation were informed by the qualitative methods of anthropology, which seek to explain how people create social orders and meanings.39 Through this process, we recognized that our respondents described everyday work practices that corresponded to the values that scholarship and political discourse often demand of accountable democratic governance.
Like any method, ours provides only a partial view of the object of analysis. Our limited number of participants constrains the range of experiences we encompassed, though we find this limitation in breadth amply balanced by the depth of discussion elicited by our wide-ranging, lengthy interviews. The extent to which our institutionally diverse respondents converged on key parts of their work leaves us confident that we have identified some central aspects of our administrative state.
Nat’l Fed’n of Indep. Bus. v. DOL (NFIB), 142 S. Ct. 661, 669 (2022) (Gorsuch, J., concurring) (quoting Antonin Scalia, A Note on the Benzene Case, Regul., July/August 1980, at 27, https://www.cato.org/sites/cato.org/files/serials/files/regulation/1980/7/v4n4-5.pdf [https://perma.cc/5ATY-QNVV]); see also Josh Blackman, Video and Transcript of Justice Alito’s Keynote Address to the Federalist Society, Volokh Conspiracy (Nov. 12, 2020, 11:18 PM), https://reason.com/volokh/2020/11/12/video-and-transcript-of-justice-alitos-keynote-address-to-the-federalist-society [https://perma.cc/8CDG-EGTL] (providing a transcript of Justice Alito’s speech in which he decries the growing “dominance of lawmaking by executive [f]iat rather than legislation”).
Critiques of the “deep state” amplify this concern for accountability and can in turn be used to argue for a unitary executive branch insulated from external constraints on the President’s power. For a recent and penetrating account of the interrelationship between the deep-state trope and claims for the unitary executive, see generally Stephen Skowronek, John A. Dearborn & Desmond King, Phantoms of a Beleaguered Republic: The Deep State and the Unitary Executive (2021).
NFIB, 142 S. Ct. at 666 (Breyer, Sotomayor & Kagan, JJ., dissenting). The literature on agency accountability is vast and largely focused on legitimating bureaucratic power and the constraints placed on it. For a powerful elaboration of the history and recent growth of critiques of the administrative state, see generally Gillian E. Metzger, The Supreme Court, 2016 Term—Foreword: 1930s Redux: The Administrative State Under Siege, 131 Harv. L. Rev. 1 (2017).
We summarize our methodology below and explain it in detail in the Methods Appendix, which discusses the research subject population, interview questions, and analysis process. We did not set out to write about agency accountability. Instead, our study focused on how agencies engage in statutory interpretation as a daily practice—how they work with their statutes to produce policy. In related work, we will present what we learned about agencies’ relationships with their statutes. But we present this Article first, because when asked how they work with statutes, officials across agencies described practices that seemed to promote aspects of accountability often attributed to elections, such as pluralistic inclusivity, reasoned deliberation, and responsiveness to situations and opinions. Although distinct from elections, these factors resonate with the ideals of republican democracy. See, e.g., Philip Pettit, Republican Freedom and Contestatory Democratization, in Democracy’s Value 163, 164-65, 180 (Ian Shapiro & Casiano Hacker-Cardón eds., 1999) (arguing that a “contestatory mode of democratization” in which private parties serve not as “authors” but as “editors[]” of the law, serves the goals of “republican freedom” understood as “non-domination,” that is, freedom from “arbitrary” power).
The “protean” concept of accountability is often used as a cover for normative debates about policy effects. See Jerry L. Mashaw, Structuring a “Dense Complexity”: Accountability and the Project of Administrative Law, 5 Issues Legal Scholarship 1, 15, 19-20 (2005). We therefore offer a fairly minimal definition of the concept, removed from any particular policy content—a concept we think should be broadly acceptable.
See Edward Rubin, The Myth of Accountability and the Anti-Administrative Impulse, 103 Mich. L. Rev. 2073, 2073 (2005); Glen Staszewski, Reason-Giving and Accountability, 93 Minn. L. Rev. 1253, 1254 (2009); see also Anya Bernstein, Agency in State Agencies, in Distributed Agency 41, 46-47 (N.J. Enfield & Paul Kockelman eds., 2017) (noting that accountability “may always depend on . . . the social conditions that render a particular attribution of responsibility felicitous”). An accountable actor may be subject to discipline for wayward—that is, not properly accounted for—actions, as well as to recognition for approved ones. Although discussions of accountability sometimes focus on the availability of sanctions, in our view it is the requirement, and the opportunity, to explain and justify positions to relevant audiences who can evaluate, affect, and potentially override them that is crucial.
See Susan Rose-Ackerman, Democracy and Executive Power: Policymaking Accountability in the US, the UK, Germany, and France 2 (2021) (“Disparate policy views are normal in a democracy. Unanimous consent is not a realistic goal for most policy choices.”); Bonnie Honig, Political Theory and the Displacement of Politics 72 (1993) (describing Friedrich Nietzsche’s argument that legal strictures do not resolve democratic contestation over values and practices but instead allow contestation to continue); see also Daniel E. Walters, The Administrative Agon: A Democratic Theory for a Conflictual Regulatory State, 132 Yale L.J. 1, 8-15 (2022) (outlining a vision of administrative agencies as a site of democratic contestation to legitimate the administrative state).
Seila L. LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183, 2204 (2020) (quoting Free Enter. Fund v. Pub. Co. Acct. Oversight Bd., 561 U.S. 477, 499 (2010)). The elections-accountability connection has also been a keystone of a well-organized and well-funded antiregulatory movement that dates back at least to the Reagan era. See generally Steven M. Teles, The Rise of the Conservative Legal Movement: The Battle for Control of the Law (2012) (tracing this history); Stephen M. Teles, Transformative Bureaucracy: Reagan’s Lawyers and the Dynamics of Political Investment, 23 Stud. Am. Pol. Dev. 61 (2009) (describing the political aims of the Reagan Department of Justice (DOJ)). This movement has emphasized the accountability exacted by elections in contradistinction to bureaucracy’s supposed remove from popular control. Staszewski, supra note 8, at 1254 (“[I]n what might be considered optimistically circular reasoning, modern public law typically presumes that elected officials are politically accountable . . . because they are selected and potentially removed from office by the voters.”).
Members of Congress also help to shape day-to-day agency behavior, but we focus in Part I on relations within the executive branch, not least because this form of political involvement in agency decision-making is far more routinized, regularized, and effective. In related work, we argue that, though day-to-day interaction with current members of Congress helps shape policymaking, agencies’ primary fidelity remains to their statutes and the regimes those statutes created—not to the enacting Congress itself, much less the current Congress. See Anya Bernstein & Cristina Rodríguez, Activating Statutes (2023) (unpublished manuscript) (on file with authors).
See, e.g., Rubin, supra note 8, at 2073. Recent scholarship has lent increasing complexity to principal-agent models. See infra note 175. However, the common conceptions of accountability as presented in legal doctrine and much legal scholarship continue to assume fairly simple and limited relationships between principals and agents. It is this simplified image that we address.
In this sense, our work contributes to an emerging account of accountability that revolves around pluralistic networks. As Francesca Bignami puts it, scholarship needs to “recast[] administrative law as an accountability network of rules and procedures through which civil servants are embedded in their liberal democratic societies,” which produces both administrative and political accountability. Francesca Bignami, From Expert Administration to Accountability Network: A New Paradigm for Comparative Administrative Law, 59 Am. J. Compar. L. 859, 861 (2011).
The substantial literature pointing to the frailties of the American electoral system both underscores this point and makes clear the system’s present-day limitations. See, e.g., Sanford Levinson, Our Undemocratic Constitution: Where the U.S. Constitution Goes Wrong (and How We the People Can Correct It), 60 Bull. Am. Acad. Arts & Scis. 31, 31-33 (2007); Ari Berman, Give Us the Ballot: The Modern Struggle for Voting Rights in America 10-12 (2015). The system’s structural minoritarianism is exacerbated by well-documented practices such as gerrymandering and voter suppression, the disenfranchisement of social subgroups (such as those with criminal records), and so on, which further undermine the possibility of electoral accountability. One would expect those committed to the electoral model to focus on eradicating these pathologies first and foremost.
See, e.g., Metzger, supra note 4, at 71 (describing schools of thought in legal scholarship arguing that the “administrative state enables the exercise of unaccountable and aggrandized executive power” because “[u]nelected bureaucrats wield a combination of de facto legislative, judicial, and executive powers outside of meaningful political or judicial constraint”); Philip Hamburger, Is Administrative Law Unlawful? 6, 19 (2014) (arguing that administrative power is “extralegal” because the Constitution authorizes only elected members of the legislature, not unelected administrators, to make rules with binding force). This view resembles the “chain-of-legitimacy model,” or Legitimationskette, of German law. Rose-Ackerman, supra note 10, at 3-4. Significant recent scholarship challenges the historical pedigree of this view of legitimacy in the United States. See, e.g., Nicholas R. Parrillo, A Critical Assessment of the Originalist Case Against Administrative Regulatory Power: New Evidence from the Federal Tax on Private Real Estate in the 1790s, 130 Yale L.J. 1288, 1302-04 (2021); Julian Davis Mortenson & Nicholas Bagley, Delegation at the Founding, 121 Colum. L. Rev. 277, 279-80 (2021); Christine Kexel Chabot, The Lost History of Delegation at the Founding, 56 Ga. L. Rev. 81, 86-88 (2021).
Scholars have drawn attention to the limits of elections. See, e.g., Rubin, supra note 8, at 2077-83; Staszewski, supra note 8, at 1269-70; David B. Spence & Frank Cross, A Public Choice Case for the Administrative State, 89 Geo. L.J. 97, 123 (2000) (“Even if voters do choose politicians who share their personal values, politicians’ policy choices are not always driven by those personal values. Rather, the relative immediacy and complexity of the electoral connection for politicians offers a great deal of opportunity for values that differ from those of the median voter to influence the policy choices of elected politicians.”); Spence & Cross, supra, at 124 (noting that politicians may cater to “particular constituencies,” play off the high or low salience of particular issues to particular groups, and generally let “self-interest, the desire for re-election, [] contaminate the policy choice”); see also Spence & Cross, supra, at 124 (“When an elected politician makes a policy choice so as to minimize the adverse electoral consequences of that choice at the next election, that decision may reflect a variety of values that are different from those that would have guided the median voter under those same circumstances.”).
Indeed, insisting on thorough presidential control may actively undermine accountability. For instance, Heidi Kitrosser has argued that the unitary executive vision allows the President “to shield or manipulate the very information” that voters need to make an informed decision. Heidi Kitrosser, The Accountable Executive, 93 Minn. L. Rev. 1741, 1743 (2009). It obscures the multiple people and processes involved in decision-making “with a single, intrinsically opaque and relatively inaccessible formal decision maker,” but still allows the President to “distance himself from unpopular decisions” by pinning the blame on “inferiors within an opaque executive branch.” Id. A model that puts all the responsibility in the single presidential node, then, undercuts accountability by impeding information flow. Id. Moreover, “empirical evidence . . . does not support the claims that independent bureaucrats advance their own interests at the expense of the commonwealth; to the contrary, greater independence may better promote the public interest.” Spence & Cross, supra note 22, at 119.
The crucial role of administration in democratic governance was recognized long ago: as Max Weber noted, “Bureaucracy inevitably accompanies modern mass democracy.” 2 Max Weber, Economy and Society: An Outline of Interpretive Sociology 983 (Guenther Roth & Claus Wittich eds., Ephraim Fischoff trans., 1978) (emphasis omitted).
See supra notes 1-2 (collecting cases); see also Cristina M. Rodríguez, The Supreme Court, 2020 Term—Foreword: Regime Change, 135 Harv. L. Rev. 1, 117-20 (2021) (arguing that the Court’s recent removal-power jurisprudence “prevents Congress from making complex trade-offs and determining how best to sustain good governance and legitimacy within the administrative state”).
This is a universal phenomenon, not something unique to administration. Cf. James S. Fishkin, The Televised Deliberative Poll: An Experiment in Democracy, 546 Annals Am. Acad. Pol. & Soc. Sci. 132, 137 (1996) (finding that participants in an information-gathering and deliberation experiment came to “new, considered judgments” over the course of deliberation).
See Michael Lipsky, Street-Level Bureaucracy and the Analysis of Urban Reform, 6 Urb. Affs. Q. 391, 392 (1971); see also Daniel E. Ho & Sam Sherman, Managing Street-Level Arbitrariness: The Evidence Base for Public Sector Quality Improvement, 13 Ann. Rev. L. & Soc. Sci. 251, 252-53 (2017) (surveying this literature); Tatjana Thelen, Larissa Vetters & Keebet von Benda-Beckmann, Stategraphy: Relational Modes, Boundary Work, and Embeddedness, in Stategraphy: Toward a Relational Anthropology of the State 1, 1-2 (Tatjana Thelen, Larissa Vetters & Keebet von Benda-Beckmann eds., 2018) (collecting chapters analyzing “how specific state constellations and boundaries emerge and are reproduced or dissolved”); Bernardo Zacka, Political Theory Rediscovers Public Administration, 25 Ann. Rev. Pol. Sci. 21, 22-24 (2022) (discussing the recent interest of political theorists in the administrative state).
We also build on studies which have taken the kind of synthetic, transsubstantive view we do here. See generally Marissa Martino Golden, What Motivates Bureaucrats?: Politics and Administration During the Reagan Years (2000) (finding the complex interplay between political appointees and civil servants to be central to policymaking in a time of political change); Joel D. Aberbach & Bert A. Rockman, In the Web of Politics: Three Decades of the U.S. Federal Executive (2000) (finding high responsiveness to legislative impetus in the administrative state).
A range of illuminating work has been done on specific agency processes. See, e.g., Daniel Carpenter, Reputation and Power: Organizational Image and Pharmaceutical Regulation at the FDA (2010) (exploring the ways in which the Food and Drug Administration (FDA) has cultivated a reputation for competence and vigilance and the impact of that organizational image on the agency’s effectiveness as a regulator); R. Shep Melnick, Regulation and the Courts: The Case of the Clean Air Act (1983) (discussing the impact of court decisions on the Environmental Protection Agency’s (EPA’s) policymaking processes); Adam B. Cox & Cristina M. Rodríguez, The President and Immigration Law (2020) (discussing the leading role of the presidency and the executive branch in the formation of immigration policy); Thomas O. McGarity, The Internal Structure of EPA Rulemaking, 54 Law & Contemp. Probs. 57 (1991) (canvassing the internal dynamics of EPA’s rulemaking process).
Our interviewees worked in presidential administrations from President George H.W. Bush to President Trump, though only a few worked in either bookend administration. The majority worked in the Obama Administration. We assume that some policymaking practices changed significantly over this long stretch of time. Particularly during the Trump Administration, some political appointees were reported to have had the explicit goal of enervating the bureaucracy. See generally David L. Noll, Administrative Sabotage, 120 Mich. L. Rev. 753 (2022) (providing several examples of this phenomenon). Political control and interagency processes surely changed in response. Still, given the density and longevity of the bureaucracy as compared with any particular presidential administration, the breadth and complexity of the structures and practices we identified seem likely to persist. And our study provides insight into the supports available for administrations who do not seek to undermine the statutory mandates of the agencies they staff.
See Anya Bernstein, Bureaucratic Speech: Language Choice and Democratic Identity in the Taipei Bureaucracy, 40 Pol. & Legal Anthropology Rev. 28, 42 (2017) (arguing that bureaucracy is “an always localized phenomenon[,] . . . a loose organizational form that maintains certain similarities over times and places, but is always situated within, and reflective of, very particular sociocultural contexts and historical trajectories, with their attendant values, beliefs, and practices”).
See Bernstein, supra note 8, at 42 (“Whether accountability is available, and what it looks like . . . depends not on some underlying relation between actor and action, but on the scaffoldings that structure the interpretation of action in particular social arenas. Understanding accountability scaffoldings in particular social realms is thus central to understanding the nature of accountability generally. Indeed, it may be more accurate not to speak of accountability generally, but only of local tropes of accountability.”).
See Clifford Geertz, “From the Native’s Point of View”: On the Nature of Anthropological Understanding, 28 Bull. Am. Acad. Arts & Scis. 26, 29 (1974) (“The trick is to figure out what the devil they think they are up to.”). We did not attempt a true ethnography of our subjects in the tradition of anthropology, but we conducted and analyzed our interviews with an “ethnographic attitude,” that is, an eye to understand how systems and processes make sense from the inside, as well as evaluating them from the outside. Anya Bernstein, Saying What the Law Is, Law & Soc. Inquiry (forthcoming) (manuscript at 1-2), https://ssrn.com/abstract=3990976 [https://perma.cc/GU55-QSWF].
We developed the codes and coded the transcripts along with a team of research assistants, which brought numerous views of the data into play at this key point. See Antony Bryant & Kathy Charmaz, Grounded Theory Research: Methods and Practices, in The Sage Handbook of Grounded Theory 1, 1 (Antony Bryant & Kathy Charmaz eds., 2007) (“[Grounded theory] method is designed to encourage researchers’ persistent interaction with their data, while remaining constantly involved with their emerging analyses . . . . The iterative process of moving back and forth between empirical data and emerging analysis makes the collected data progressively more focused and the analysis successively more theoretical.”).
See Anne Rawls, Harold Garfinkel, in The Blackwell Companion to Major Contemporary Social Theorists 122, 122-23 (George Ritzer ed., 2003) (describing “ethnomethodology” as investigating the “shared methods” that people have “for achieving social order that they use to mutually construct the meaningful orderliness of social situations,” on the understanding that “the meaningful, patterned, and orderly character of everyday life is something people must work constantly to achieve”).
Bernstein, supra note 36 (manuscript at 3) (“Analyzing an ethnographic object involves recognizing it as a social product: shot through with cultural forces and exerting social effects. It also involves taking multiple perspectives on the object of analysis to explain how the social phenomenon makes sense to the people who engage in it, exploring how that sense-making fits into larger contexts, and illuminating some of what it leaves out or glosses over. Ethnographic inquiry seeks to approach its objects from many directions at once. We view our objects from the inside, but we also take perspectives that diverge from those the objects themselves suggest, and we stay on the lookout for implicit underpinnings and effects.”).